ABSTRACT

Purpose:

it has been shown that serum prolactin level usually rises after epileptic seizures. Therefore measurement of postictal serum prolactin is recommended as a useful test for the differentiation of epileptic seizures from pseudoseizures in suspected events. However, there are studies that report serum prolactin level elevation also after nonepileptic atacks. in this study, we investigated postictal changes of serum prolactin values in epileptic and pseudoepileptic seizures.

Material and methods:

Ali of the patients were young males who had been hospitalized for seizure observation. Baseline serum prolactin level was measured at the next morning after hospitalisation. Postictal meisure was made just after seizure occured in ten minutes. Differentiation of siezure type (true epileptic or pseudoepileptic) was made by direct observation and video-camera records, independently from prolactin meisures.

Results:

There were 47 patients in epilepsy and 69 in pseudoepilepsy group. Postictal serum prolactin levels were elevated significantly in both groups when compared with their baselines, but more prominent in epilepsy group. Postictal levels of two groups were also statistically different (higher in epilepsy group) while baselines were not. The sensitivity, spesifity, positive and negative predictive value were calculated separately for 1.5, 2, 2.5, 3, 3.5 and 4-fold relative increases (postictal/baseline).

Conclusion:

As regards the lower relative increases (especially lower than two-folds), the spesifity and positive predictive value for epileptic seizure are not satisfactory. But the sensitivity and negative predictive value are high (For 1.5, sensitivity 91%, spesifity 65%, positive predictive value 64% and negative predictive value is 92%). If we use the higher relative increases (three or two-folds) as criterion, the spesifity and positive predictive value are high but not 100% Furthermore the sensitivity and negative predictive value are low (For 4, sensitivity 36%, spesifity 97%, positive predictive value 89% and negative predictive value is 69%). We emphasise that however postictal serum prolactin measurement may be useful in some conditions, should be interpreted cautiously according to the rate of relative increase.